The Department for Education has begun briefing schools on the use of generative AI for tutoring and lesson planning, framing the move as a productivity gain. The department's own modelling, seen by Business Fortitude, shows it is principally a response to a teacher shortage no recruitment campaign has fixed.
England is short 17,400 secondary teachers against the modelled requirement for the 2026-27 academic year, according to internal DfE forecasting circulated to senior officials in March. The shortfall is concentrated in mathematics, computing, modern foreign languages, and design and technology. The same forecast assumes a 6 per cent attrition rate among teachers in their first five years, broadly stable from the previous year.
That gap is what the AI guidance is designed to absorb.
The new framework, published this week, recommends three tools for evaluation: a maths-tutoring system from Eedi, a lesson-planning assistant from a London startup whose name has not been publicly confirmed, and Microsoft Copilot for Education (NASDAQ: MSFT). Schools will not be required to adopt any of them, but headteachers in target regions will receive a £4,000 implementation grant from the autumn term.
A productivity story written backwards
The framing is unusual. Most government productivity narratives begin with a target and add tooling to reach it. This one begins with a workforce gap and reverse-engineers a productivity claim. The DfE's own impact assessment estimates that the recommended tools could absorb 4 to 7 hours per week of teacher workload, freeing capacity equivalent to roughly 11,000 full-time teachers. That is two-thirds of the modelled shortfall.
If we hit our recruitment numbers we don't need this. We aren't hitting our numbers, so we do.
Senior DfE official, speaking on background
The honest read is that AI is being deployed not because it is the best pedagogical tool but because the alternative, a 17,000-teacher recruitment surge, is not in prospect. The Treasury has been clear with departmental colleagues that there is no headroom for a multi-year salary uplift of the scale required. Schools Week first reported the recruitment shortfall in February.
What boards of trust academies need to ask
For the boards of multi-academy trusts, which now educate 56 per cent of secondary pupils in England according to the latest DfE statistics, three questions will dominate the autumn governance cycle.
First, what is the baseline of teacher time the AI tools are meant to displace? Lesson planning is a useful place to start because it is well-defined; feedback on student work is harder, and the displacement claim there is weaker.
Second, what is the data residency posture? Several of the tools recommended by the framework process pupil work outside the UK. The DfE has not yet issued a detailed data-protection annexe; the Information Commissioner's Office is reportedly preparing one for the autumn.
Third, what is the parental-communication strategy? Trusts that have already piloted Copilot in one or two academies have found that parents will accept AI in lesson planning more readily than in marking, irrespective of the tool's technical capability.
The DfE has not commented publicly on the modelling. A spokesperson described the policy as evidence-based and additive to recruitment, which is neither a confirmation nor a denial of the workforce framing.
What remains true is that the policy is doing two jobs at once. It is encouraging classroom use of AI, which is a worthwhile goal in its own right. And it is providing political cover for a teacher-shortage problem that the government does not currently have a plan to solve. Operators in the UK education sector should price both jobs into how they read the framework.
